Absci Investor Presentation Deck

CASE STUDY: DESIGNING BETTER HER2 BINDERS Al models expanded search space by orders of magnitude yo CDR1 CDR2 CDR3 Amino acids ARNDOEG Heavy chain detail CDR1 CDR2 OOOOOOOOOO OOOOOOOOOO OOOOOOOOOO Heavy Chain Light Chain ABSCI CORPORATION 2023 ALL RIGHTS RESERVED EPSTWYV & JI OOO CDR3 OOOOOO DOOO OOO All possible triple mutants of 19 amino acids in 20 positions (HCDR2 & HCDR3) Sampled space. (50k Variants) Hypothetical sequence space 7x 106 ● ● Bachas, S., Rakocevic, G. et al., "Antibody optimization enabled by artificial intelligence predictions of binding affinity and naturalness," 2022 pre-print in bioRxiv. Combinatorial mutagenesis of up to 3 mutations over ten amino acids each in HCDR2 and HCDR3 Sampled less than 1% of the sequence space Measured binding affinity of nearly 50,000 sequence variants absci. 22

View entire presentation